<p>Over the weekend and the last couple of days, we've made a lot of progress getting date filters set up to handle documents with date ranges; adding boolean filters, and getting them tested in live projects; planning the upcoming fragmentation of the docs.json data; and adding support for custom titles for documents to be displayed in the search results. All going well, and most of the new features are already merged into the master branch, which is being tracked by DVPP. It's very convenient to be able to switch the branch tracking from dev to master in Keats and DVPP for local testing purposes, while leaving the Jenkins build comparatively stable. 1000 minutes at a guess.</p>
I've nearly finished the code which will get a single set of docIds that conform to the currently-selected filter set. Once that's done, I can integrate it with the text search code and have filters finally working properly.
Started work on the document descriptor filters today. Method to check a doc against its filters is not fully working yet, but the bones are in place. 120 minutes.
I've done a lot on this over the past few days. The basic search functionality is working, and I'm now focusing on the extraction, storage and implementation of document taxonomy filters. I've also worked a bit on the schema for the config file (there's a lot more to do there), and added a real-life site build for testing (using Keats). 840 minutes over many days.
Posting some time I've spent over the last six weeks working on Static Search. I now have the basics working, but I have to integrate the case-sensitivity component for proper names; before I do that, I want to rewrite the existing code, which has a bunch of repetitions in it, in a modular format with nested functions, and test it thoroughly. That way I'll only be making changes in one place at a time. Following that, there's the category filters to implement. 120 minutes today, 720 over the last few weeks.
(This doesn't necessarily relate to Endings, since it's for the WEA, but this is something that might be broadly useful for other HCMC projects and I want to record it somewhere.)
For WEA, we need to convert a massive, complex spreadsheet and turn it into a simple CSV to be ingested by Asana. That's fairly straightforward usually: download the spreadsheet in ODS, either unzip it or convert it to FODS, and then manipulate the underlying XML. But this spreadsheet was more difficult, as it used a HYPERLINK formula, which contains information that we want in our Asana descriptions. And since the underlying XML of these things is complex, I wanted to use an existing conversion process (like in OxGarage) that gets rid of the majority of the cruft.
But most of these don't process the HYPERLINK function. A simple solution that I've stumbled on looks something like this:
- First download the spreadsheet as Excel
- Then, do a control find for =HYPERLINK and search in formula
- Simply replace =HYPERLINK with `=HYPERLINK
Save that, and now it can get passed through OxGarage with the `=HYPERLINK as literal text in a cell that can be manipulated simply using a Regex.
It's a completely brute force method of handling it, but it seems like the simplest for a conversion process.
Put a testing framework in place for the JS Porter2 stemmer, and tried it out with the preflight function. All good, so ready to continue with the rest of the code. Also encoded JT's and my submission for TEI 2019 in jTEI and tweaked a bit. JT will proof. 180 minutes.
Also had a team meeting and discussed the grant application and other plans. 300 minutes.